capgo-native-builds

Pass

Audited by Gen Agent Trust Hub on Nov 13, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill frequently uses npx @capgo/cli@latest to execute the latest version of the Capgo CLI directly from the NPM registry. This is a standard deployment method for the vendor's tools.
  • [COMMAND_EXECUTION]: The skill relies on executing various shell commands via the @capgo/cli to automate native cloud builds, credential management, and store uploads.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to read and interpret local project configuration files, which presents a potential surface for indirect prompt injection if those files contain malicious instructions. However, the skill includes explicit warnings for the agent to verify actions and protect sensitive data.
  • Ingestion points: Reads capacitor.config.*, android/app/build.gradle, and package.json to extract application metadata (documented in SKILL.md and references/request-options.md).
  • Boundary markers: No specific delimiters or "ignore previous instructions" markers are applied to the content of these project files.
  • Capability inventory: The skill can perform network operations (build requests, uploads), file writes (exporting credentials to .p8 or .env files), and credential management (build credentials save).
  • Sanitization: The instructions do not specify sanitization or validation logic for the data parsed from project configuration files before using it in CLI commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 13, 2026, 12:05 AM